SnapCatch
Rajeswari Balaji, R. Karthik, Anuprabha Kiran, Obuli Prasad J. · Advances in computational intelligence and robotics book series · 2025
Covert Timing Channels (CTCs), which exploit transmission delays to steal private data, are increasingly dangerous on digital platforms. Traditional detection methods often lack flexibility and harm service quality. This study introduces CryptoSleuthNet, a novel framework combining Multimodal Generative AI with Blockchain to detect CTCs without disrupting normal traffic. By transforming network traffic into artificial visual data, the method enables intuitive detection. Neural networks, Naive Bayes, and Elliptic Curve Cryptography (ECC) classifiers analyze these datasets, with ECC achieving the highest accuracy (80%). Blockchain smart contracts boost forensic reliability and transparency by ensuring tamper-proof tracking and verified message validity. This hybrid approach enhances detection, traceability, and trust—critical in combating data exfiltration and deepfakes. By merging AI and decentralized security, CryptoSleuthNet marks a major advance in safeguarding media integrity and detecting hidden threats in dynamic, content-rich ecosystems.